-->

Keep up with all of the essential KM news with a FREE subscription to KMWorld magazine. Find out more and subscribe today!

Conversational Agents and AI Transformation: Q&A with Rob Ryan, Head of Business Value and Thought Leadership, Workgrid

Ojala: Thinking in terms of AI, GenAI, agentic AI, what do you think is going to come next for AI assistants?

Ryan: For our space, conversational AI, we are moving to a phase where AI doesn't just assist with tasks or deliver news and notifications, but takes that next step. Right now, most AI will handle straightforward tasks, but we're heading to a future where decisions will execute on their own. It's a shift from just simply delivering answers to a future with a handling of complex queries requiring backend reasoning. AI agents allow for really iterative reasoning to take place—instead of throwing a list of possible answers at you, it'll go deeper, potentially even asking questions to itself, "Did I fully answer the question of the user?"

We could ask a question such as, "Which of my contracts are expiring this year?" And the agent might scan through contact tracks, identify new ones that are set to expire, giving you a clear answer. It could even, potentially, proactively start to nudge you on those as well, and then provide you with those next steps and guidance. So the future is all about AI seamlessly connecting to those multiple systems, applications, in taking a personalized approach to provide a cohesive response and removing that ambiguity from next steps. So really having that true partner to get things done.

Ojala: How far along are we do you think we are with agentic AI?

Ryan: It's early days. So there's a lot of companies who are certainly attempting to claim that they have the full agent suites out there ready to go. To be quite honest, everyone's at the same starting point.

Ojala: One of the issues with knowledge management has been, and to some extent, remains is knowledge ensconced in silos. How well can the agents circumvent that and go from one silo to another?

Ryan: Part of that we're allowing agents to have access to the right information, systems, and portals, and also defining what those portals happen to be for that agent to begin to crawl and take those next necessary steps. That's why you see a combination between both the LLMs that are grounded in vertical knowledge, let's say if it's for IT or for insurance And then of course, providing it to the right systems and access to ensure that it has the ability to reason, take those next steps, and then find that information and extract it for the employee.

So it is certainly exciting days, but at the end of the day, if you look towards more of the agentic piece, that's certainly going to require data cleanup, proper knowledge management, and at the end of the day, garbage in, garbage out.

Ojala: What advice would you give to organizations for implementing AI in their organizations, from day-to-day applications to actually scaling AI projects?

Ryan: AI adoption as a whole is very fragmented depending on the organization, depending on the market business vertical. Right now, there's a lot of pilots going on, early projects that are focused on content generation summarization like ChatGPT or Copilot, the real game changer of AI is embedding into those deeper systems of operations, not just surface level use cases of generate a blog, generate news. AI should be evolving with your organization, your culture, and your appetite for risk. That's where Workgrid certainly comes in. Our platform is built to meet you wherever you happen to be on that journey.

We offer ready to go use case templates out of the gate that are secure within closed environments, so your teams can get comfortable with AI. That can be for content generation or very specific business applications. We seamlessly connect to third party applications and helpi deliver notifications and nudges, based upon events that might be happening within their operational change. So the key takeaway here is: Don't rush. It has to be right for you. The crawl, walk, run approach hasn't changed. Get your teams comfortable, really build from there, and AI should feel like a natural extension and expression of day-to-day operations, not just another tool you're trying to shoehorn into the organization.

KMWorld Covers
Free
for qualified subscribers
Subscribe Now Current Issue Past Issues